pic: Drive gearbox shorter than 2" Tall?!



Hey everyone!
(skip to the bottom for CAD links)
Hope you like my latest creation! In poking around 1678’s CAD I noticed their drive gearboxes were really cool. I really liked the overall design and the form of their drive gearboxes (What happened to ballshiftsohard??!), so after cadding up my own gearbox in the same style, I thought it was cool but would be cautious to use it on a competition robot in a typical game because it couldn’t shift and only had 2 CIMs. From there, I thought if it had 3 CIMS and could stay similar in size/simplicity it would be something more suited for my team to try on our offseason bot and maybe a season bot. My first step was research, and what did I find? Chak already did that. And it’s really cool. And I don’t understand what’s going on in half of it.

Chak’s awesome design is linked here, for anyone interested:

But I had been pretty sick for a week and really needed a design project to kill another day or two before I could go back to work, and I thought with the rise of 775pro drivetrain gearboxes it might be cool to design one of my own. My primary goal was (beyond the obvious light, functional, as small as possible, etc) to design a gearbox that could completely fit under the top surface of a tube, without the need for a gearbox slot in the bellypan.

I ended up with this gearbox, and sitting nicely at 1.9" tall (stolen from my largest gear’s OD from vex) it’s definitely the coolest gearbox I’ve designed.

Some variability in the sprockets could vary the gearbox speeds across different designs, but a realistic sample is a 16 and a 22 tooth sprocket from VP or whoever end up with a 14.98 adjusted speed from JVN.

My next iteration will be a very similar gearbox with everything (but the motors ofc) inside of a 2x2 along (imagine the plates replaced with a two by two) with sprockets, and I guess we will then have a design for gearbox and chain in tube!

Anyways, Here is the Onshape link:

And a google drive link with a STEP:

As always, questions and feedback are appreciated!

Neat design. Love the compact size, reminds me of custom drive gearboxes my team used to make way back in 2006 and prior.

My only real criticisms are:

  1. It’s not very weight efficient (since you need 2x the gears, though I understand this is a factor of the size)
  2. Using chains (susceptible to slop and loss of tension over time)

I feel like there’s probably a way to avoid using chains by replacing the outer sprockets with a gear, using a larger idler gear, and moving the idler and output shaft gears to the opposite end of the gearbox (closer to the “camera” in your picture). I would worry that not only would chains be a huge pain to maintain, but would also be more prone to failures.

You may have imported this into Onshape, but if you didn’t (and for everyone else), you should check out the multi-part part studio feature in Onshape. It allows you to create multiple parts all in the same studio and have them reference each other. In my opinion, its the most powerful part of Onshape aside from the cloud.

You should also check out the mate connector tutorial. Once you understand how Onshape uses mate connectors, it’ll save you a lot of time over the “traditional” mating systems in SW, Autodesk, etc.

As for the gearbox itself, do you need the chain reductions if you turn this into an 8WD chassis?

I definitely didn’t copy this into Onshape. As you can see in some of my other creations on here I do often put multiple parts in part studios. However, for this gearbox a lot of it involved figuring things out as I go (this is my first multi-stage gearbox), and I kept the parts separate so that they didn’t mess with each other through referenced geometry and stuff that I typically use.

As for Mate stuff, I’m definitely stuck on some old Solidworks habits, so thanks for sharing the link as it’s a lot more helpful than the brief explanations for some of the specific mates I don’t really get yet.

As for your comment about the chain reduction, I’m not quite sure what you mean. Getting rid of the run altogether? Running the chain just not on a reduction? A little more background would be appreciated.

Right now you’re bringing all your power into a central wheel for what I assume is a 6wd chassis. What if you had two output shafts from your gearbox to power the two center wheels of an 8wd chassis? Maybe this would allow you to eliminate the chain runs in the gearbox.

Check out 118 from 2016 for a good example.

Or, along those lines, have 4 separate gearboxes (2 on each side of the robot) each with 2x 775Pros. There wouldn’t really be a much of a weight penalty since you’re already using 2x the gears.

If you want all of the wheels to have power from all 4 motors, you’ll still need to chain the gearboxes together. And if your chain breaks, you’ll likely start smoking motors if you get into to a pushing match and you won’t bale able to drive at all instead of just missing one wheel. At that point, I don’t see the benefit to the added complexity.

IMO, I would argue it’s simpler to use two gearboxes per side than to have a gearbox with two output shafts. You would have to connect two driven shafts with chains in either case, so I was simply comparing the differences in complexity between a single gearbox with two output shafts vs two independent gearboxes that could presumably be made with far fewer parts.

That said, I think we’re all overlooking the simplest solution here:

Move all the motors to one end of the gearbox and put the output shaft on the other end rather than centered between the motors. You would need to add another 32dp idler gear, but you could eliminate 4 sprockets, 2 chain runs, and one of the small 20dp gears, which should more than make up for it.

Now you’re placing some of the 775Pro output shafts in bending. Might work. Might fail.

I’d love to see some destructive testing. ‘will it bend’?

[Can’t delete this post, it’s attached to a picture, missed the conversation thread]]]


As opposed to what? They’re in bending with pinions on them driving a single gear already.

My bad, I misread your post and thought you were suggesting 4 gearboxes per side (i.e. one per wheel). I agree that two separate gearboxes on the middle wheels each with a chain to the other and to the outside wheel wouldn’t be much more complicated than one gearbox with two output shafts.

A possible problem with moving the output shaft all the way to one end is that all of the force for all 4 motors is going through the last motor pinion. They can probably take it, but I’d be careful.

We used to do this back in the day with CIM motors and never had any issues (albeit, the pinion gears were slightly larger), but it’s a fair critique.

In other news… I got board and CADed a gearbox today…
http://i.imgur.com/nJEjpME.jpg](https://i.imgur.com/PTtFkzU.jpg)

It’s actually pretty hard to get a gearbox with the correct reduction in a less-than 2in tall space, as it turns out.

I’ve seen 254 bend shafts like that just passing 2 CIM’s worth of power through a single pinion. If you gear sufficiently slowly it’s not a problem, but why risk it when you can just design things not to need to? In any case, I don’t like using idlers for the wasted resources they take up.
I would not run more than a single 775’s power through a pinion. Given that, I don’t think there are many ways to improve this gearbox, other than maybe making it easier to have different output ratios and (maybe?) adding a spacer closer to the center of the gearbox for stiffness. Great work! If I have the resources I’ll try doing this in the offseason.

Belts could work pretty easily here in place of the chain

I used chain here cause the availible differences in tooth count is much closer together for small sprockets than belts. Makes it wasy to swap between combinations of 16, 18, and 22 teeth sprockets in the gearbox for different output shaft speeds.

You could add an idler, though that would be more complex and less efficient. Though, with such small chain runs here you may have issues with swapping because of incorrect c-c anyways.